Hooked on A.I.: will recreational fishing catch the machine learning bug?
A few friends, a handful of rods and a leisurely weekend on the water: Recreational fishing doesn’t seem like it should have environmental implications. But it does.
That’s because there are 13.2 million saltwater fishers in the United States. Even accounting for the ones that got away, that’s a high enough catch to put some populations, such as red snapper in the Gulf of Mexico, at risk.
It’s a challenge that regulators tasked with tackling overfishing and ensuring ecosystems remain sustainable have wrestled with for some time. But without clear data indicating how many fish are being caught, it’s been hard to set, or enforce, scientifically sustainable limits.
Now, help is at hand. A new monitoring system that uses cameras and machine learning to observe boats and help calculate catches, is undergoing trials in the U.S. and overseas. The SmartPass system combines photo technology and artificial intelligence to identify the type and total number of vessels that pass in and out of ports. This, paired with calculations of dockside catch, will give regulators a more accurate picture of what’s being caught, in what quantities and where.
“This technology has the potential to fill the biggest gap in fisheries management there is out there,” says Environmental Defense Fund’s manager of oceans technology Dr. Chris Cusack, who helped devise the SmartPass system. “Good science-based fisheries management cannot happen without good data. It feeds policy decisions, fights poaching and helps maintain stocks.”
Currently most monitoring of recreational fisheries in the U.S. is done on a boat-by-boat basis, in-person or through phone and email surveys. It’s labor intensive, expensive and, because of the huge number of vessels on the water, only samples a fraction of the fleet. SmartPass uses cameras mounted at ‘pinch points’ such as port entrances or narrow passes to capture images of every boat that passes in or out. The photos are then processed by a machine learning algorithm that identifies the type and number of boats. Combining that information with the size of catches brought ashore, gives a more precise picture of what’s going on beneath the sea. The speed and completeness of the information allows fishery managers to respond rapidly and with a greater degree of certainty to changing fishing patterns and populations. It can help regulators set sensible limits, adjust limits when stocks are low or extend the season when the data shows that populations are in good health.
The Oregon Department of Fish and Wildlife has been using video monitoring to capture recreational fishing data for some time. It’s now looking to machine learning capabilities of SmartPass to make further progress.
“The more accurate and efficient we can be, the better we can manage the fisheries for recreational anglers and conservation needs,” said Eric Schindler, project leader of the Ocean Sampling Project for ODFW.
“Oregon has been a trailblazer in innovative fisheries management and we’re excited to be partnering with the Department of Fish and Wildlife on developing ways to collect more timely and robust data.” says EDF’s Sepp Haukebo, co-developer of the SmartPass.
Thriving fish populations help ensure the sustainability of fisheries and the health of entire ocean ecosystems. But it’s not just life under water that benefits. Every season, thousands of anglers travel to the Gulf of Mexico to fish, spending their dollars on charter boats, restaurants, lodging and more — at precisely the time when other tourism undergoes its annual seasonal dip. Keeping fish stocks healthy, keeps local businesses going too.
“Snowbirds like myself pump millions of dollars into the economy,” says Minnesota-based recreational angler, Jeff Senarighi, who spends up to three months a year fishing in the Gulf. “If science tells us we can harvest a certain amount and still build the fisheries back up, that’s what we should be doing.”
SmartPass can also benefit fisheries further afield. Globally, small boat fishing is estimated to account for more than half the fish taken from the sea. As oceans warm and fish populations move, SmartPass can supply critical information about fluctuations in stocks that will help the world’s fishers continue to feed the 3 billion people globally who depend on fish for protein.
“With climate change we’ll see changes in currents and weather patterns and that affects fish populations,” Haukebo adds. “Those changes could upend a lot of our assumptions about what good fisheries management looks like.”
In Indonesia’s Lampung province, thousands of fishers scour the shallow waters for blue swimming crab every day. But with little to no data on the number of boats, nor how many crabs are being caught, it is feared the fishery could be under threat. Fishers and the government are looking for ways ensure the health of the fishery and secure livelihoods for local communities now and in the future. It is hoped SmartPass could be part of the solution.
“Blue swimming crab is the third biggest export for Indonesia,” says fisheries enumerator, Victor Pandapotan Malau. “The value is very high. With good data collection, we can know whether the stock is still sufficient or not.”
